Rank-order stability of relationship satisfaction: A meta-analysis of longitudinal studies

Satisfaction with a romantic relationship often changes over time, and individuals differ in how satisfied they are in their relationship. However, no systematic review is available regarding the stability of individual differences in relationship satisfaction. Therefore, this meta-analysis synthesizes the available longitudinal data on rank-order stability of relationship satisfaction, as a function of age and relationship duration. Analyses were based on 148 samples including 153,396 participants reporting on their relationship over time. Mean age associated with the effect sizes ranged from 19 to 71 years, and mean relationship duration from 3 months to 46 years. On average, individual differences in relationship satisfaction were highly stable over time (r = .76, corrected for attenuation due to measurement error and based on an average time lag of 2.30 years). Rank-order stability varied systematically as a function of age, increasing from young to late adulthood with a slight decline during middle adulthood. Rank-order stability also varied as a function of relationship duration, increasing over the course of the relationship with a slight decline around 20 years of relationship duration. Moderator analyses suggested that relationship transitions shortly before Time 1 and sample type explained variance in rank-order stability. However, except for these two moderators, the pattern of findings was robust across all characteristics tested. In sum, this meta-analysis indicates that relationship satisfaction is a relatively stable construct, with lower stabilities in young adulthood and in the first years after beginning a relationship. This knowledge may stimulate future research on developmental processes within romantic relationships.